This paper extends organizational research on distrust to explain the effects of organizational misconduct on continued market participation after a fraud. Iargue that social relations between fraud victims and perpetrators insulate against the formation and diffusion of distrust. Variation in market participation after a fraud occurs because victims who belong to the perpetrator’s social group are more likely to attribute blame to the organization that committed the fraud, while victims from social out-groups are more likely to generalize blame to the perpetrator’s social group and subsequently avoid other organizations and institutions governed by its members. The empirical setting is Kenya’s ethnically diverse Nairobi Securities Exchange (NSE), from which the country’s largest stock brokerage was expelled in 2008 after defrauding one-quarter of its100,000 clients. Analysis of NSE data on trades for victimized and non-victimized investors, as well as surnames that identify investors’ ethnicities,shows that clients from the same ethnic group as leaders of the corrupt broker-age are more likely than clients from rival ethnic groups to continue to participate in the market after a fraud and more likely to choose another intermediary operated by members of their ethnic group. But results also show that some victims invest more after the fraud and use less-trustworthy intermediaries to do so. This suggests that integrity-based trust may be a less effective coordination device in arm’s-length market transactions than in interpersonal transactions. A key implication is that misconduct does not simply reduce average participation in a market; it changes the market’s composition by filtering out diverse social groups that are more likely to demand stronger governance standards.